Complement anaphylatoxin C5a contributes to hemodialysis-associated thrombosis

Abstract

Thrombosis is a common complication of end-stage renal disease, particularly in patients on hemodialysis. Although substantial progress has been made in preventing thrombotic complications in various other groups of patients, the mechanisms of thrombosis during hemodialysis require clarification. In this report, we demonstrate that complement activation triggered by hemodialysis biomaterials, and the subsequent generation of the complement anaphylatoxin C5a, results in the expression of functionally active tissue factor (TF) in peripheral blood neutrophils. Because TF is a key initiator of coagulation in vivo, we postulate that the recurring complement activation that occurs during long-term hemodialysis contributes to thrombosis in dialyzed end-stage renal disease patients. Furthermore, we found that complement contributed to the induction of granulocyte colony-stimulating factor, which has been implicated in the pathogenesis of thrombosis in patients treated with the recombinant form of this molecule. Importantly, the inhibition of complement activation attenuated the TF expression and granulocyte colony-stimulating factor induction in blood passing through a hemodialysis circuit, suggesting that the complement system could become a new therapeutic target for preventing thrombosis in patients with chronic renal failure who are maintained on hemodialysis.

Figure 1

ESRD patients' sera stimulate the production of functionally active TF in peripheral blood leukocytes from healthy donors. (A-B) The mPT values of supernatants of normal PMNs (A) and PBMCs (B) incubated with PBS (bar 1 in panels A and B), serum from healthy donors (HI, bar 2 in panels A and B), or ESRD sera isolated before (bar 3 in panels A and B; predialysis [Pred]) or during hemodialysis at various time points (bars 4-8 in panels A and B); mPT values of supernatants of PMNs (A) and PBMCs (B) either incubated with APS serum (bar 10 in panels A and B) or incubated with ESRD sera collected between 15 and 30 minutes after the beginning of hemodialysis and treated with TF monoclonal antibodies (bar 11 in panels A and B). Data are representative of 10 independent experiments (mean mPT values ± SD). A paired t test was applied to examine statistical significance: *P < .05; **P < .001. (C) The relative expression of TF mRNA in normal PMNs cultured in the presence of PBS or ESRD patients' sera isolated before or during hemodialysis. Data are representative of 6 independent experiments (mean fold expression [2^−DDCt analysis] ± DC_t SD). DDCt indicates Delta Delta Cycle threshold; and DCt, Delta Cycle threshold. The Wilcoxon matched-pairs test was used to assess statistical significance; *P < .01. Horizontal lines above data bars indicate statistical significance.

Figure 2

TF is expressed in the peripheral leukocytes of ESRD patients. (A) The relative expression of TF or asTF mRNA in PMNs (i) and PBMCs (ii) obtained from healthy donors (HI, bar 1) or ESRD patients before (predialysis [Pred], bar 2) and during hemodialysis (bars 3, 4). Data are representative of 6 independent experiments (mean fold expression ± DC_t SD). The Wilcoxon matched-pairs and Mann-Whitney tests were used to assess statistical significance: *P < .05. (B) The induction of TF expression in PMNs (i) and PBMCs (ii) isolated from ESRD patients before (bar 2) and during hemodialysis (bars 3, 4). The results are presented as ratios of the mean fluorescence intensities of the cells from patients to the cells from healthy donors (bar 1), stained with TF monoclonal antibody, and analyzed by flow cytometry. Data are representative of one experiment (mean ± SD of 4 healthy donors and 6 ESRD patients). The Mann-Whitney and Wilcoxon matched-pairs tests were used to assess statistical significance: *P < .05. (C) TF expression in PMNs isolated from healthy donors (1-HI) or ESRD patients before (2-predialysis [Pred]) and during hemodialysis (lanes 3, 4). One representative Western blot analysis of 4 independent experiments is shown. Horizontal lines above data bars indicate statistical significance.

Figure 3

Procoagulant properties of ESRD sera are complement-dependent. (A) The amounts of soluble TCC in plasma from healthy donors (HI, bar 1) or ESRD patients before dialysis (predialysis [Pred], bar 2), and during (bars 3-6) hemodialysis. Data are representative of one experiment (mean ± SD of 4 persons per group). The Mann-Whitney test and paired t test were applied to assess statistical significance: *P < .05. (B-C) The expression of C5a receptor on the surface of PMNs (B) or in permeabilized cells (C) from healthy donors (HI), ESRD patients before dialysis (predialysis [Pred]), and during hemodialysis (5, 30, 60, and 120 minutes). The results are presented as mean fluorescent intensity (MFI) ratios of the cells stained with C5aR monoclonal antibodies to the cells stained with the isotype control. Data are representative of one experiment (mean ± SD of 4 healthy donors and 3 ESRD patients). The unpaired and paired t tests were used to assess statistical significance: *P < .05. (D) The mPT values of supernatants of normal PMNs treated with sera from healthy donors (HI, bar 1), ESRD patients before (predialysis [Pred], bar 2), and during (bar 3) hemodialysis or preincubated with C5a (bar 4) or C3a (bar 5) receptor antagonists and then treated with sera obtained from ESRD patients during hemodialysis. Data are representative of 6 independent experiments (mean ± SD). The Wilcoxon matched-pairs test was used to assess statistical significance: *P < .001. (E) The relative expression of TF mRNA in normal PMNs treated with sera obtained from ESRD patients during hemodialysis or preincubated with C5aR antagonist and then treated with sera from the same patients. Data are representative of 2 independent experiments (mean ± SD of 3 patients). The Wilcoxon matched-pairs test was used to assess statistical significance: P < .01. Horizontal lines above data bars indicate statistical significance.

Figure 4

Hemodialysis filter fibers induce complement activation and enhance TF-dependent procoagulant properties in normal and ESRD sera. (A) The amounts of C3 cleavage products in plasma isolated from normal nontreated blood (bar 1), nontreated blood incubated in 37°C for 60 minutes (bar 2), blood treated with filter fibers (bar 3), or blood treated with filter fibers in the presence of compstatin analog (bar 4) or control peptide (bar 5). Data are representative of 4 independent experiments (mean ± SD). The paired t test was used to assess statistical significance: *P < .05. (B) The mPT values of supernatants of normal PMNs incubated with sera from healthy donors (HI, bar 1), sera activated with filter fibers for various periods of time (bars 2-5), or with filter fiber-activated sera (for 60 minutes) and treated with TF monoclonal antibodies (bar 6). Data are representative of 6 independent experiments (mean ± SD). The Wilcoxon matched-pairs test was used to assess statistical significance: *P < .01. (C) The mPT values of supernatants of normal PMNs incubated with sera from healthy donors (HI, bar 1), ESRD patients before hemodialysis (predialysis [Pred], bar 2), filter fiber-activated sera from healthy donors or patients (bars 3 and 6, respectively), sera from healthy donors or patients incubated with these fibers in the presence of compstatin analog (bars 4 and 7, respectively) or pretreated with C5aR antagonist and then incubated with fiber-activated sera from healthy donors or patients (bars 5 and 8, respectively). Data are representative of 6 independent experiments (mean ± SD). The Wilcoxon matched-pairs test was used to assess statistical significance: *P < .01. (D) The induction of TF expression in normal PMNs incubated with sera from healthy donors (bar 1), fiber-activated sera (bar 2), or sera treated with fibers in the presence of compstatin analog (bar 3). The results are presented as ratios of the mean fluorescence intensities (MFI) of the cells incubated with sera treated with fibers in the presence or absence of compstatin to the cells incubated with sera from healthy donors, stained with TF monoclonal antibodies, and analyzed by flow cytometry. Data are representative of 6 independent experiments (mean ± SD). The Wilcoxon matched-pairs test was used to assess statistical significance: *P < .05. (E) TF expression in normal PMNs incubated with PBS (1), APS sera (2), fiber-treated sera in the presence of compstatin analog (3), or fiber-activated (4) or untreated sera (5). One representative Western blot analysis of 4 independent experiments is shown. Horizontal lines above data bars indicate statistical significance.

Figure 5

Compstatin analogs attenuate the hemodialysis-associated activation of complement and neutrophils and up-regulation of TF in these cells. Normal blood was recirculated through a hemodialysis simulation system for 2 hours at 37°C. (A) The amounts of C3 cleavage products in plasma isolated from the blood before the beginning of this recirculation (predialysis [Pred]) and at various time points during hemodialysis simulation in the presence of compstatin analogs or control peptide. (B) The induction of TF and (C) CD11b expression in PMNs isolated from the blood before (predialysis [Pred]) and during hemodialysis simulation. The results are presented as ratios of the mean fluorescence intensities (MFI) of the cells stained with TF monoclonal antibodies to the cells stained with the isotype control and analyzed by flow cytometry. Data are representative of 4 independent experiments in panel A and 3 in panels B and C (mean ± SEM). The 2-way analysis of variance test was applied to assess statistical significance of the difference between treatment of blood with compstatin analogs (▴) and the control peptide (■): (A) P < .001, (B) P < .034, and (C) P < .021.

Figure 6

Complement regulates cytokine production and/or release during a hemodialysis simulation. The amounts of (A) IFN-γ, (B) IL-1RA, and (C) G-CSF during hemodialysis simulation. Data are representative of 3 independent experiments (mean ± SEM). The 2-way analysis of variance test was used to assess statistical significance of differences in the amounts of cytokines after treatment of the blood with compstatin analogs (▴) or the control peptide (■): (A) P < .018, (B) P < .030, and (C) P > .008.